Tài liệu Báo cáo khoa học: Calcium-binding to lens bB2- and bA3-crystallins suggests that all b-crystallins are calcium-binding proteins pptx

Calcium-binding to lens bB2- and bA3-crystallins suggests
that all b-crystallins are calcium-binding proteins
Maroor K. Jobby and Yogendra Sharma
Centre for Cellular and Molecular Biology (CCMB), Hyderabad, India
Crystallins are abundant proteins found in the eye lens
of vertebrates that belong to two superfamilies named
as a-crystallins and bc-crystallins [1]. a-Crystallins are
known to play an important role as molecular chaper-
one [2]. On the other hand, bc-crystallins are thought
to play structural role in the mammalian eye lens.
Their nonstructural functions, which appear to be very
important, have not been elucidated [3].
b-Crystallins from vertebrate eye lens are a group of
seven proteins broadly classified into four acidic
(bA1 ⁄ A3, bA2 and bA4) and three basic b-crystallins
(bB1, bB2, and bB3). b-Crystallins have high sequence
similarity and identity [4]. Acidic b-crystallins have
both N- and C-terminal extensions, whereas basic
b-crystallins have only N-terminal extensions. All
b-crystallins have four Greek key motifs organized into
two crystallin domains. In this respect, b-crystallins are
similar to c-crystallins, which also have a similar
domain organization and structure [5,6]. The major
difference between the b- and c-crystallins is their
oligomeric state. c-Crystallins are monomeric, whereas
b-crystallins exist as dimers to octamers in solution [7].
b- and c-crystallins are the prototype and founding
members of the bc-crystallin superfamily [8,9].
Keywords
bA3-crystallin; bB2-crystallin; bc-crystallins;
calcium-binding crystallin; Greek key motif

Correspondence
Y. Sharma, Centre for Cellular and
Molecular Biology (CCMB), Uppal Road,
Hyderabad 500 007, India
Fax: +91 40 2716 0591
Tel: +91 40 2716 0222
E-mail:
(Received 28 April 2007, revised 11 June
2007, accepted 14 June 2007)
doi:10.1111/j.1742-4658.2007.05941.x
Crystallins are the major proteins of a mammalian eye lens. The topologic-
ally similar eye lens proteins, b- and c-crystallins, are the prototype and
founding members of the bc-crystallin superfamily. bc-Crystallins have
until recently been regarded as structural proteins. However, the calcium-
binding properties of a few members and the potential role of bc-crystallins
in fertility are being investigated. Because the calcium-binding elements of
other member proteins, such as spherulin 3a, are not present in bB2-crys-
tallin and other bc-crystallins from fish and mammalian genomes, it was
argued that lens bc-crystallins should not bind calcium. In order to probe
whether b-crystallins can bind calcium, we selected one basic (bB2) and
one acidic (bA3) b-crystallin for calcium-binding studies. Using calcium-
binding assays such as
45
Ca overlay, terbium binding, Stains-All and
isothermal titration calorimetry, we established that both bB2- and
bA3-crystallin bind calcium with moderate affinity. There was no signifi-
cant change in their conformation upon binding calcium as monitored by
fluorescence and circular dichroism spectroscopy. However,
15
N-

1
H hetero-
nuclear single quantum correlation NMR spectroscopy revealed that amide
environment of several residues underwent changes indicating calcium
ligation. With the corroboration of calcium-binding to bB2- and bA3-crys-
tallins, we suggest that all b-crystallins bind calcium. Our results have
important implications for understanding the calcium-related cataracto-
genesis and maintenance of ionic homeostasis in the lens.
Abbreviations
AIM1, protein absent in melanoma 1; HSQC, heteronuclear single quantum correlation; ITC, isothermal titration calorimetry; PDB, protein
databank; TCEP, Tris(2-carboxyethyl) phosphine hydrochloride.
FEBS Journal 274 (2007) 4135–4147 ª 2007 The Authors Journal compilation ª 2007 FEBS 4135
bc-Crystallin superfamily consists of members from
various taxa having the characteristic crystallin-type
Greek key motifs [8,10]. Some well studied members
of the superfamily are Protein S [11,12], spherulin 3a
[8,13], protein absent in melanoma 1 (AIM1) [14,15],
geodin [16], ciona crystallin [17], yersinia crystallin [18]
and cargo proteins from Tetrahymena [19].
Except for some conserved residues present at cru-
cial positions, there is not much sequence similarity
among the diverse proteins of the bc-crystallin super-
family. Recently, it has been proposed that these
bc-crystallins might play unknown and unconceived
noncrystallin roles [3]. These ‘noncrystallin roles’ have
not been elucidated to date. We are interested in
understanding the nonstructural functions of bc-crys-
tallins. Previously, we reported that c-crystallins bind
calcium [20], and therefore, might be involved in main-
taining calcium homeostasis in lens. Recently, bB2-

crystallin has been implicated in the subfertility of mice
expressing mutant bB2-crystallin [21]. Some proteins of
the superfamily, Protein S, spherulin 3a [10], bc-crys-
tallin domains of AIM1 [14,15], yersinia crystallin [18],
geodin [22] and ciona crystallin [17] are known to bind
calcium ions.
However, the binding of calcium to b-crystallins is
inconclusive and highly debatable [10,23], even though
the aggregated form of b-crystallins, b
H
-crystallin, iso-
lated from bovine lens homogenate was shown to bind
calcium [24,25]. Sequence D ⁄ NXXS, which is involved
in calcium-binding in Protein S, spherulin 3a and in an
invertebrate ciona crystallin [17,23,26], is not conserved
in vertebrate lens b-crystallins. Furthermore, the cal-
cium-ligating side chains and the backbone conforma-
tion of spherulin 3a are structurally not conserved in
bB2-crystallin [23]. Accordingly, it has been argued
that b-crystallins should not bind calcium. In the light
of these contradictory observations, it is important to
investigate whether b-crystallins from vertebrate lenses
bind calcium or not.
In this context, to establish calcium-binding to the
individual b-crystallins, we have selected a basic (bB2-
crystallin) and an acidic (bA3-crystallin) subunit as
representative members of b-crystallins. Using number
of assays for proving specificity of calcium-binding, we
have conclusively demonstrated that both acidic and
basic b-crystallins bind calcium with varying affinity,

thus suggesting that all b-crystallins would bind
calcium. Calcium-binding does not influence protein
conformation, a property exhibited by some of the
calcium-binding members of the bc-crystallin super-
family [14,15,20]. Based on our results, together with
the published data on calcium-binding to a few
other members, we suggest that calcium-binding is a
prevalent property of the bc-crystallin superfamily.
Demonstration of calcium-binding to b-crystallins
would fill an important and missing link in our exist-
ing knowledge about bc-crystallins as calcium-binding
proteins and understanding their function in maintain-
ing calcium homeostasis in the lens, which is impli-
cated in cataracts.
Results and Discussion
Selection of b-crystallins
The sequence alignment of seven b-crystallins [four aci-
dic (A1–A4) and three basic (B1–B3) crystallins] is
shown in Fig. 1. There is 45–60% sequence identity
between different b-crystallins [4]. We have selected
one acidic (bA1 ⁄ A3-crystallin) and one basic (bB2-
crystallin) subunit as representatives of all b-crystallins
for probing the calcium-binding properties. We have
selected bB2-crystallin because it is the major crystallin
among all b-crystallins and its 3D structure is known
[27]. bA1- and bA3-crystallins are identical in sequence
except for N-terminal extension of 17 amino acids in
bA3-crystallin. Moreover, these b-crystallins have been
widely studied for structural properties and hetero-
and homo-domain interactions with each other as well

as with other b-crystallin subunits [7,28]. These pro-
teins have been predicted not to bind calcium
[10,17,23]. We believe that studying these two b-crys-
tallins would provide an insight into the calcium-bind-
ing properties of all b-crystallins.
Overexpression and purification
Bovine bB2- and bA3-crystallin were cloned in expres-
sion vector and overexpressed in Escherichia coli as
recombinant proteins. Proteins were purified using a
combination of chromatographic methods. The purity
of each batch of protein was confirmed by examining
the samples on SDS ⁄ PAGE (supplementary Fig. S1).
Protein solutions were treated with Chelex-100 for
removing divalent ions and used as fresh as possible
for further calcium-binding studies, otherwise the pro-
teins were stored frozen at )80 °C.
Calcium-binding to bB2- and bA3-crystallins
Because there is no known motif for calcium-binding
in bB2- and bA3-crystallins, it was therefore necessary
that calcium-binding should be assayed by several spe-
cific methods. We used well-known calcium probes,
Stains-All (Sigma-Aldrich, St Louis, MO, USA) and
terbium binding to assess the calcium-binding. We also
All lens b-crystallins are calcium-binding proteins M. K. Jobby and Y. Sharma
4136 FEBS Journal 274 (2007) 4135–4147 ª 2007 The Authors Journal compilation ª 2007 FEBS
used direct calcium-binding on membrane using
45
Ca. The binding constants and other thermodynamic
parameters were determined using isothermal titration
calorimetry.

Probing calcium-binding by Stains-All assay
Calcium-binding to bB2- and bA3-crystallins was eval-
uated by calcium probe Stains-All, a carbocyanine dye
[29]. The dye binds the recombinant bA3- and bB2-
crystallins and induces a strong J band at 660 nm
(Fig. 2). The intensity of the circular dichroic band
decreases upon addition of calcium ions because cal-
cium displaces the dye bound to calcium-binding sites
of the protein. Other proteins of this superfamily,
namely c-crystallin [20] and AIM1-g1 [15] also induce
the J band of the dye indicating similarity in the
microenvironment of the dye-binding site [30]. Calcium
saturated proteins exhibited no binding to Stains-All
dye, suggesting higher affinity of the cation for the
calcium-binding site than the dye. Calcium displaced
Stains-All to a lesser extent from bA3-crystallin than
from bB2-crystallin, indicating lower affinity of
calcium for the former compared to the latter.
Fig. 1. Sequence alignment and putative calcium-binding sites: Amino acid sequences of six bovine b-crystallins were aligned using Multialin.
Putative calcium-binding residues are indicated by asterisks. Green line marks the Greek key motif.
M. K. Jobby and Y. Sharma All lens b-crystallins are calcium-binding proteins
FEBS Journal 274 (2007) 4135–4147 ª 2007 The Authors Journal compilation ª 2007 FEBS 4137
Probing calcium-binding by terbium
We also probed calcium-binding using another calcium
probe, terbium. The ionic radius of terbium is similar
to that of calcium, thus making it an ideal choice for
use as a calcium mimic probe [31]. Terbium ions bind
to the calcium-binding sites in proteins and induce
luminescence peaks at 492 nm and 547 nm via energy
transfer from Trp and Tyr residues [32]. Terbium binds

to bB2- and bA3-crystallins and induces luminescence
peaks at 492 and 547 nm (Fig. 3). The enhanced
luminescence of terbium in the presence of these crys-
tallins indicates that Tyr and Trp residues are in the
vicinity of the calcium-binding site. The sequence of
b-crystallins has several Tyr and Trp residues distributed
around the putative calcium-binding residues of both
crystallins, resulting in the observed increase in inten-
sity (Fig. 1). Similar results were observed with the D2
domain of yersinia crystallin [18], which also had a
Trp residue near the second calcium-binding site. We
also carried out a terbium–calcium competition assay.
bB2- and bA3-crystallins presaturated with calcium
showed increased fluorescence intensity upon adding
increasing concentrations of terbium, which indicated
that terbium displaced the bound calcium. This is
expected because terbium ions have a higher affinity
A
B
Fig. 2. Stains-All binding to (A) bB2- and (B) bA3-crystallins: 100 lg
of either bB2- or bA3-crystallin protein was added to Stains-All dye
in 2 m
M Mops ⁄ NaOH (pH 7.2) and 30% ethylene glycol and CD
spectra were recorded from 400–700 nm. (A) Calcium was added
to a final concentration of 25, 300 and 5300 l
M. (B) calcium was
added to a final concentration of 0.5, 1.5 and 8.5 m
M. Arrows indi-
cate increasing concentrations of calcium.
A

B
Fig. 3. Terbium binding to b-crystallins: (A) 7.68 lM of bB2- and
(B) 22.68 l
M of bA3-crystallin were excited at 285 nm and emission
spectra recorded from 300–560 nm. Terbium was added to a final
concentration of 0, 5, 25, 45, 65, 85, 300, 700 l
M to bA3-crystallin
and 0, 15, 35, 55, 85, 500, 1200 and 3200 l
M to bB2-crystallin.
Inset shows the region from 480–555 nm. Arrows indicate increas-
ing concentrations of terbium.
All lens b-crystallins are calcium-binding proteins M. K. Jobby and Y. Sharma
4138 FEBS Journal 274 (2007) 4135–4147 ª 2007 The Authors Journal compilation ª 2007 FEBS
than calcium for calcium-binding sites in the protein
due to the higher positive charge of terbium than
calcium [31].
Calcium-binding by
45
Ca overlay method
Calcium-binding was also demonstrated by direct
45
Ca-binding using the membrane overlay method [33].
This simple and direct assay has been widely used to
ascertain the cation binding to calcium-binding pro-
teins. Both b-crystallins immobilized on nitrocellulose
membrane bound calcium, whereas the negative con-
trol BSA did not show any binding (Fig. 4). The buffer
used for this assay contained MgCl
2,
another divalent

cation that usually competes for calcium-binding sites
in proteins, despite which we observed positive signal
from bA3- and bB2-crystallin immobilized on the
membrane. This demonstrates the specificity of these
proteins for calcium unlike EF-hand proteins, which
bind both calcium and magnesium. In control experi-
ments, we have carried out
45
Ca-binding to these crys-
tallins in the presence of cold CaCl
2
and found that
the signal was abolished (data not shown).
Calcium-binding by isothermal titration calorimetry
The cation-binding constants of both crystallins were
determined by isothermal titration calorimetry (ITC)
measurements. Calcium-binding to bB2-crystallin is an
exothermic reaction (Fig. 5A). The integrated heats of
injection of calcium titration to bB2-crystallin best
fitted to a sequential binding model with four sites.
By varying the initialization parameters of the fitting
procedure, it was determined that the fit was stable
and no other model and parameter set could provide a
Fig. 4.
45
Ca overlay: 50 lg of BSA, bB2- and bA3-crystallins were
spotted on a nitrocellulose membrane. The processed membrane
was exposed to imaging plate before scanning in a phosphor
imager (Fuji FLA-3000).
AB

Fig. 5. Isothermal titration calorimetry: (A) calcium-binding isotherm of bB2-crystallin. (B) Terbium binding isotherm of bA3-crystallin. The best
fit to four-site sequential binding model is shown in the lower panels.
M. K. Jobby and Y. Sharma All lens b-crystallins are calcium-binding proteins
FEBS Journal 274 (2007) 4135–4147 ª 2007 The Authors Journal compilation ª 2007 FEBS 4139
better fit. The dissociation constants of calcium-bind-
ing to bB2-crystallin range from 0.16 mm to 83 lm
(Table 1). These results reveal the presence of four
calcium-binding sites with moderate to low affinity.
Stains-All and terbium-binding studies indicated
that bA3-crystallin has relatively lower affinity for
the cation than bB2-crystallin. Calcium-binding to
bA3-crystallin studied by ITC resulted in poor signal
as expected and, thus, this method was unsuitable
for determining the binding constants of calcium to
bA3-crystallin (data not shown). We, therefore, carried
out terbium binding to this crystallin by ITC and
determined the binding constant for the calcium
mimic probe. Terbium is believed to bind strongly to
calcium-binding sites of proteins compared to calcium
due to its higher charge ratio than calcium, even
though both ions have similar ionic radii [31]. The
dissociation constants of terbium-binding to bA3-
crystallin range from 2.7 mm to 40 lm (Table 1). The
low affinity might explain the nonsaturating nature
of binding thermogram (Fig. 5B). Calcium is thus
likely to bind to bA3-crystallin with lower affinity than
terbium.
The above results using specific assays for calcium-
binding, suggest that both bB2- and bA3-crystallins
bind calcium with moderate affinity. We have observed

that these proteins lose the calcium-binding ability
upon storage and specific precautions, such as the use
of freshly prepared protein, are required to perform
calcium-binding experiments.
Effect of calcium on protein conformation
We further studied the effect of calcium-binding on the
conformation, stability and hydrodynamic radii of
these crystallins using fluorescence spectroscopy, CD
spectroscopy, differential scanning calorimetry, analyt-
ical gel filtration and dynamic light scattering.
Trp fluorescence emission spectra
Trp fluorescence emission spectrum is an important
tool in probing the microenvironment of Trp residues
in proteins. We used this to probe the changes in the
polarity of Trp residues upon calcium-binding. bA3-
and bB2-crystallins exhibited emission maxima at 342
and 333 nm, respectively, indicating that Trp residues
in both proteins are in nonpolar environment (Fig. 6).
Calcium-binding does not induce any significant
changes in the emission spectra of both crystallins;
however, only minor changes were seen in case of
bA3-crystallin (Fig. 6B).
Table 1. Binding constants and the enthalpy change of calcium-
and terbium-binding to bB2- and bA3-crystallins. K, dissociation con-
stant (M); DH, enthalpy change of binding (kcalÆmol
)1
).
Parameters
bB2-Crystallin
(calcium-binding)

bA3-Crystallin
(terbium binding)
K
1
(2.15 ± 1.3) · 10
)4
(1.08 ± 0.08) · 10
)4
K
2
(1.65 ± 0.98) · 10
)4
(1.46 ± 0.09) · 10
)4
K
3
(8.33 ± 7.63) · 10
)5
(4.03 ± 0.3) · 10
)5
K
4
(5.71 ± 4.89) · 10
)4
(2.72 ± 0.13) · 10
)3
DH
1
2.75 ± 0.65 2.76 ± 0.07
DH

2
4.0 ± 1.0 2.15 ± 0.23
DH
3
)2.4 ± 0.86 )0.98 ± 0.26
DH
4
0.61 ± 0.28 13.5 ± 0.34
A
B
Fig. 6. Fluorescence spectroscopy: 7 lM of each protein was exci-
ted at 295 nm and emission recorded from 300–450 nm. Calcium
was added to the desired concentration and incubated for 5 min
before recording the emission spectra. (A) Emission spectra of b B2-
crystallin: Final concentration of calcium added was 0, 0.5, 2, 12,
30, 100, 1000 l
M (B) Emission spectra of bA3-crystallin: final con-
centration of calcium added was 0, 0.5, 1,6, 24, 80, 1000, 2000,
3000 l
M. Arrows indicate an increasing concentration of calcium.
All lens b-crystallins are calcium-binding proteins M. K. Jobby and Y. Sharma
4140 FEBS Journal 274 (2007) 4135–4147 ª 2007 The Authors Journal compilation ª 2007 FEBS
Far- and near-UV CD spectroscopy
The native state of the recombinant proteins as well as
structural changes upon calcium-binding were monit-
ored by far- and near-UV CD spectroscopy (Fig. 7).
Far-UV CD spectra of both crystallins have a minima
around 218–220 nm characteristic of b-sheet conforma-
tion. There is a slight change in the spectra in the
region below 200 nm upon addition of calcium; how-

ever, secondary structure fractions of apo and holo
forms calculated using the program cdnn [34] indica-
ted no significant changes in both the proteins.
The near-UV CD spectra of bB2- and bA3-crystal-
lins are dominated by a broad band in the 255–285 nm
region, indicating the contribution from aromatic
amino acids and Cys (there are 5 Trp, 9 Tyr, 8 Phe
and 2 Cys in bB2-crystallin and 9 Trp, 11 Tyr, 8 Phe
and 8 Cys in bA3-crystallin) (Fig. 8). There is no
significant change in the near-UV CD spectra of both
proteins upon titration with calcium, corroborating
our results of far-UV CD and Trp fluorescence spectro-
scopy.
2D NMR spectroscopy
Each crosspeak in the
15
N-
1
H heteronuclear single
quantum correlation (HSQC) spectrum of a protein
represents an amide bond of amino acids in the
A
B
Fig. 7. Far-UV CD spectroscopy: (A) 0.71 mgÆmL
)1
of bB2-crystallin
and (B) 2.1 mgÆmL
)1
of bA3-crystallin in 10 mM Tris-Cl (pH 7.5) and
30 m

M KCl was used for recording the far-UV CD spectra. Calcium
aliquots were added from a standard stock solution to a final con-
centration of 0, 0.1, 1 and 10 m
M to bB2-crystallin and 0, 0.5, 1
and 5 m
M to bA3-crystallin.
A
B
Fig. 8. Near-UV CD spectroscopy: (A) 1.1 mgÆmL
)1
of bB2- and
(B) 0.65 mgÆmL
)1
of bA3-crystallin was used for recording the far-
UV CD spectra. Calcium was added from a standard stock solution
to a final concentration of 0, 0.1, 0.5, 1.5 and 3.5 m
M each to either
bB2- or bA3-crystallin.
M. K. Jobby and Y. Sharma All lens b-crystallins are calcium-binding proteins
FEBS Journal 274 (2007) 4135–4147 ª 2007 The Authors Journal compilation ª 2007 FEBS 4141
protein. Perturbation of these crosspeaks upon
ligand-binding is an indication of changes in the
microenvironment of that residue. Sensitivity enhanced
2D [
15
N-
1
H] HSQC spectra were recorded. We used
this technique to determine the changes in
15

N-
1
H
HSQC spectra of the bB2-crystallin upon calcium-
binding (Fig. 9). Three spectra corresponding to apo,
half-saturated and saturated proteins have been over-
lapped for comparison. Some of the residues marked
in the box underwent changes in peak intensity and
position in the 2D
15
N-
1
H HSQC spectrum upon cal-
cium titration, suggesting calcium ligation. The large
size of the protein due to known homodimerization
and higher oligomer formation with increasing protein
concentration makes it difficult to carry out the neces-
sary 3D NMR experiments for assignment of residues
of this protein [35]. Also, a number of structures for
bB2-crystallin are available in protein databank (PDB)
structures solved by X-ray crystallography [6,27,36,37].
We also carried out the differential scanning calori-
metry, analytical gel filtration and dynamic light
scattering of the apo and holo forms of bA3- and
bB2-crystallins. There was no significant change in the
stability and hydrodynamic radius of the both forms
of proteins (data not shown).
These properties are similar to the results on few
other proteins of this superfamily such as c-crystallin
[20], AIM1-g1 [15], AIM1-g5 [14] and D2 domain of

yersinia crystallin [18], in which calcium-binding does
not cause significant changes in protein conformation.
This might suit to their function as calcium buffers
because they are not expected to transduce signals as
calcium sensors by conformational change upon cal-
cium-binding.
All b-crystallins are calcium-binding proteins
We have for the first time evaluated the calcium-bind-
ing properties of two widely studied representative
proteins of b-crystallins, bB2- and bA3-crystallin,
both by direct (
45
Ca-binding to protein on membrane
and by ITC) and methods using calcium-mimic
probes (terbium and Stains-All binding). Our results
conclusively demonstrate that both proteins bind cal-
cium with moderate affinity with no change in their
conformation, stability and hydrodynamic radii. Pro-
teins with moderate to low affinity for calcium are
also known, such as calsequestrin (with a dissociation
constant of approximately 1 mm) [38] and calreticulin
[39] belonging to the EF-hand superfamily. There is
high sequence similarity in all b-crystallins, and we
therefore suggest that all seven b-crystallins would
bind calcium.
Putative calcium-binding sites
Each Greek key motif of spherulin 3a and Protein S
contains a D ⁄ ND ⁄ NXXSS sequence element at the
loop between c–d strands, and the elements in two
motifs combine to form two symmetrical calcium-

binding sites in each crystallin domain [23,26]. This
sequence element is not exactly present in b-crystallins,
which could explain the comparatively moderate affin-
ity of these proteins as shown by our data. It has been
proposed that similar calcium-binding sites are also pre-
sent in the c-crystallins [20]. A peptide corresponding
to the third Greek key motif of c-crystallin was shown
to bind calcium whereas mutation of binding residues
abolished binding, suggesting that the motif is the mini-
mal entity required for calcium ligation [20]. The first
Greek key motif of bA3 ⁄A1-crystallins has the sequence
signature ‘DNVRS’, similar to the ‘D ⁄ NXXS’ sequence
of microbial crystallins, whereas others are diverse
(Fig. 1). Based on the comparison with Protein S and
spherulin 3a, we suggest that homologous residues in
bA3- and bB2-crystallins, would participate in calcium
ligation, as indicated in Fig. 1. We used 3D coordinates
of bB2-crystallin (PDB id 1BLB) to identify the puta-
tive calcium-binding site via the webfeature interface
[40] (supplementary Fig. S2). It will be of great interest
to define this binding motif more precisely by detailed
structural analyses from the diverse members of this
superfamily, particularly from vertebrate homologues.
Low levels of contaminating calcium ions are usually
found in laboratory solutions. Although the crystal
structures of bB2-, bB1- and c-crystallins have been
solved, calcium ion was not noticed in their solved
structures [6,41,42]. This could be due to several tech-
nical reasons. However, the most probable reasons are
the acidic pH inconducive for calcium-binding, the use

of calcium chelating phosphate buffer or protein modi-
fication during the long course of incubation resulting
in loss of calcium-binding ability. The prolonged time
required for crystallization may result in loss of the
labile, moderate to low affinity cation-binding ability
of these proteins. In vitro, we have observed that puri-
fied protein looses its calcium-binding ability upon
storage. We encountered difficulties in carrying out
ITC of several batches of bB2-crystallin, which were
not used fresh after purification. As seen in supple-
mentary Fig. S3, the signal was abolished to a large
extent and extraction of any meaningful binding
parameters was difficult. Such problems are not un-
usual and have been observed in the case of several
other calcium-binding proteins. We have also shown
previously that, despite the absence of a clear and
divergent D⁄ ND⁄ NXXSS sequence, c-crystallin and
All lens b-crystallins are calcium-binding proteins M. K. Jobby and Y. Sharma
4142 FEBS Journal 274 (2007) 4135–4147 ª 2007 The Authors Journal compilation ª 2007 FEBS
5
1
2
6
3
4
12
4
65
3
10 9 8 7 F2 [ppm]

7.8 7.7 7.6 7.5 7.4 F2 [ppm]
7.64 7.62
7.60
7.58 F2 [ppm]
8.35 8.30 8.25 8.20
F2 [ppm]
105.0105.5106.0107.0 106.5 F1 [ppm] 123.8 123.6 123.4 123.2 F1 [ppm] 113 112 111 F1 [ppm]
7.2 7.1 7.0 6.9
6.8 6.7
6.6
F2 [ppm]
8.05 8.00 7.95 7.90 7.85 7.80
F2 [ppm]
7.70 7.65
7.60 7.55
7.50
F2 [ppm]
115.5116.5 116.0 F1 [ppm] 129.5 129.0 128.5 128.0 127.5 F1 [ppm] 113 112 111 110 F1 [ppm]
125 120
115 110 F1 [ppm]
Fig. 9. 2D
15
N-
1
H HSQC spectra. The figure represents the overlap of apo, half-saturated and calcium-saturated (green, purple and red
colored contours, respectively) HSQC spectra of
15
N-labelled bB2-crystallin. Boxes in the lower panel are magnified for ease of visualization.
M. K. Jobby and Y. Sharma All lens b-crystallins are calcium-binding proteins
FEBS Journal 274 (2007) 4135–4147 ª 2007 The Authors Journal compilation ª 2007 FEBS 4143

AIM1-g1 bind calcium with affinity equivalent to
microbial crystallins [15,20].
Implications of calcium-binding to crystallins in
cataract ) a noncrystallin function
It has been known for a long time that abnormal levels
of free calcium are deleterious for the transparency of
the lens [43,44]. The mechanisms and components
involved such as sensors, buffers and modulators for
maintaining calcium homeostasis in the lens are not
known. Electron tomographic studies [45] have indica-
ted that most of the calcium in lens is bound to the
targets in fiber cell cytoplasm, with very little bound to
phospholipids near the membranes. They have sugges-
ted the presence of proteins as calcium buffer in lens
fiber cells. The moderate millimolar affinity and high
capacity calcium-binding of b-crystallins owing to their
high concentration in the lens indicate their potential
role in calcium sequestration. In other words, these
calcium-binding crystallins appeared to have been
recruited for this specialized function in the lens
during evolution. However, the physiological relevance
of calcium-binding to lens-crystallins remains to be
experimentally established. Earlier studies have linked
bB2-crystallin expression in extra-lenticular tissues to
calcium dependent stress management [46–48]. Recently,
mice harboring Philly mutation in bB2-crystallin were
found to be subfertile [21]. These studies implicate the
importance of bB2-crystallin expression for normal
physiological functions in nonlenticular tissues.
In conclusion, our data demonstrate that all b-crys-

tallins are moderate affinity calcium-binding proteins.
These results add one more calcium-binding protein to
a growing list of bc-crystallin superfamily. Our work
lays a strong foundation for the identification and
study of more proteins for calcium-binding properties
of this understudied superfamily.
Experimental procedures
Materials
All restriction enzymes and molecular biology enzymes were
from New England Biolabs Ltd (Hitchin, UK). Fine biochem-
icals were from Sigma-Aldrich, Calbiochem (Nottingham,
UK) or SRL Fine Chemicals, Mumbai, India. Plastic wares
were obtained from Tarsons Industries, Kolkata, India.
Cloning and overexpression
Cloning and overexpression of bovine bB2-crystallin has
been described previously [49]. PCR amplified bA3-crystal-
lin gene from the cDNA of bovine lens epithelial cells was
ligated to pBSK cloning vector and the insert was released
using NdeI and BamHI restriction enzymes. The insert with
cohesive ends was ligated to NdeI and BamHI digested
pET-21a using T4 DNA ligase (New England Biolabs) fol-
lowed by transformation to E. coli to select for positive
clones. The positive plasmids were sequenced to confirm
the insert sequence.
pET-21a-A3 construct was transformed to expression
host E. coli BL 21(DE3). The strain was grown in terrific
broth to mid log phase at 37 °C. When the A
600
was
between 0.6 and 1.0, isopropyl thio-b-d-galactoside was

added to the final concentration of 1 mm to induce protein
overexpression. The cultures were harvested after 3 h and
cell pellet was stored at )80 ° C.
Purification
Recombinant bB2-crystallin was purified using hydrophobic
interaction chromatography as described earlier [49]. bA3-
Crystallin was purified using anion exchanger Q-Sepharose
FF (GE Life Sciences, Piscataway, NJ, USA) using a
modified method of steady state elution [50]. The E. coli
cell pellet containing overexpressed bA3-crystallin was lysed
by ultrasonication in 50 mm Tris-Cl (pH 7.0) containing
1mm EDTA, 5 mm dithiothreitol and 5 mm phenyl-
methanesulfonyl fluoride. The clarified cell lysate was
loaded on a Q-Sepharose FF column equilibrated in 50 mm
Tris-Cl (pH 7.0) and 1 mm EDTA. Under these conditions,
bA3-crystallin does not bind to the resin. The eluate was
collected and again passed through the same column. After
two passages through the column, the protein was further
purified on a Sephadex G-75 (GE Life Sciences) column
equilibrated in 50 mm Tris-Cl (pH 7.5) containing 100 mm
KCl and 1 mm dithiothreitol. Fractions containing the pure
protein were collected and buffer exchanged with Chelex-
treated buffer to remove calcium. Proteins were either used
fresh or stored in plasticwares at )80 °C after quantitating
by absorption at 280 nm.
Stains-All binding assay
The calcium mimic dye, Stains-All, was used to probe the
calcium-binding properties of bB2- and bA3-crystallins as
described previously [29]. Briefly, 100 lg protein was mixed
with the 100 lm dye solution made in 2 mm Mops ⁄ NaOH

(pH 7.2) containing 30% ethylene glycol, and incubated for
5 min. CD spectra were then recorded between 400 and
700 nm with a 1 cm pathlength cell.
Terbium binding
Terbium-binding to both b-crystallins was monitored on a
Hitachi F-4500 spectrofluorimeter (Hitachi Corp, Tokyo,
All lens b-crystallins are calcium-binding proteins M. K. Jobby and Y. Sharma
4144 FEBS Journal 274 (2007) 4135–4147 ª 2007 The Authors Journal compilation ª 2007 FEBS
Japan). The excitation wavelength was 285 nm with band-
passes of 5 nm for excitation and emission. The buffer used
was 20 mm Tris-Cl (pH 7.5) containing 100 mm KCl.
Increasing concentrations of terbium chloride from a stock
solution (10 mm) were added to the protein solution in the
cuvette and incubated for 5 min before recording the spec-
tra from 300–560 nm.
45
Ca overlay assay
Calcium-binding to bB2- and bA3-crystallins was evaluated
by
45
Ca membrane overlay method originally described by
Maruyama et al. [33]. Proteins (50 lg each) were spotted
onto a nitrocellulose membrane using a dot-blot apparatus.
The membrane was washed with a solution containing
10 mm imidazole-HCl (pH 6.8), 60 mm KCl, 5 mm MgCl
2
and then incubated for 15 min at 25 °C in the same buffer
containing 1 lCiÆmL
)1
of

45
Ca (New England Nuclear,
Boston, MA, USA). The membrane was then rinsed twice
in 45% ethanol, dried and signal was read with a Phos-
phorimager (Fuji Bas-3000, Stamford, CT, USA).
Fluorescence spectroscopy
Fluorescence emission spectra were recorded on a Hitachi
F-4500 spectrofluorimeter. The cuvettes were soaked in
10 mm EDTA solution, rinsed with Chelex-100 treated
MQ-water (Millipore, Bedford, MA, USA) and dried
before use. The buffer used was 10 mm Tris-Cl (pH 7.5)
containing 20 mm KCl. The spectra were recorded in the
correct spectrum mode of the instrument using excitation
and emission band passes of 5 nm.
CD spectroscopy
Far- and near-UV CD spectra of both crystallins were
recorded at room temperature, on a Jasco-715 (Jasco Inc.,
Tokyo, Japan) spectropolarimeter using 0.01 cm and 1 cm
path length cuvettes, respectively. The buffer used was
10 mm Tris-Cl (pH 7.5) containing 30 mm KCl. Secondary
structure fractions from far-UV CD spectra were calculated
using cdnn based on neural networks [34].
Isothermal titration calorimetry
Calcium- and terbium-binding isotherms for bB2- and bA3-
crystallins were determined using a Microcal VP-ITC (Mic-
roCal Inc., Northampton, MA, USA). Freshly prepared
bB2-crystallin was used at a concentration of 341 lm in
10 mm Tris-Cl (pH 7.5) containing 50 m m KCl and 0.2 mm
TCEP [Tris(2-carboxyethyl) phosphine hydrochloride]. The
ligand CaCl

2
was prepared in the same buffer at a concen-
tration of 20 mm. The titration was carried out at 25 °C
using 57 injections of 4 lL each. Similarly, freshly prepared
bA3-crystallin in 20 mm Hepes ⁄ NaOH (pH 7.0), 100 m m
KCl and 0.2 mm TCEP at a concentration of 265 lm was
used in the sample cell at 20 °C. The ligand terbium chlor-
ide (10 mm) in the same buffer was loaded in the syringe
and a total of 62 injections were made. The first 13 injec-
tions were of 4 lL each and the rest of 5 lL each. The
integrated heat of each injection was used for fitting to
binding models using the program microcal origin 7.0
(Microcal Inc., Northampton, MA, USA) after subtraction
with the appropriate buffer blank.
NMR spectroscopy
15
NH
4
Cl (Cambridge Isotopes, Cambridge, MA, USA) was
used to label the recombinant bB2-crystallin overexpressed
in M9 minimal media using the protocol of Marley et al.
[51]. NMR experiments were carried out on a Bruker Avan-
ce II 600 MHz Ultrashield high resolution NMR spectro-
meter (Bruker, Ettlingen, Germany) equipped with a pulsed
field gradient unit and a triple resonance probe with act-
ively shielded Z-gradient. Sensitivity enhanced 2D [
15
N-
1
H]

HSQC spectra of the protein sample (300 lm, pH 7.5,
25 °C) were recorded. Spectra were processed using topspin
software (Brucker).
Acknowledgements
This work was supported by a grant from the Depart-
ment of Science and Technology (DST), Government of
India. MKJ is the recipient of a senior research fellow-
ship from the CSIR, Government of India. ITC facility
is supported by a Welcome Trust Senior Research the
fellowship to Dr R. Sankaranarayanan. We acknowledge
Dr Anant Patel for help with the NMR spectroscopy.
References
1 Wistow GJ & Piatigorsky J (1988) Lens crystallins: the
evolution and expression of proteins for a highly spe-
cialized tissue. Annu Rev Biochem 57, 479–504.
2 Horwitz J (1992) Alpha-crystallin can function as a
molecular chaperone. Proc Natl Acad Sci USA 89,
10449–10453.
3 Bhat SP (2004) Transparency and non-refractive func-
tions of crystallins ) a proposal. Exp Eye Res 79, 809–816.
4 Berbers GA, Hoekman WA, Bloemendal H, de Jong
WW, Kleinschmidt T & Braunitzer G (1984) Homology
between the primary structures of the major bovine
beta-crystallin chains. Eur J Biochem 139, 467–479.
5 Den Dunnen JT, Moormann RJ & Schoenmakers JG
(1985) Rat lens beta-crystallins are internally duplicated
and homologous to gamma-crystallins. Biochim Biophys
Acta 824, 295–303.
M. K. Jobby and Y. Sharma All lens b-crystallins are calcium-binding proteins
FEBS Journal 274 (2007) 4135–4147 ª 2007 The Authors Journal compilation ª 2007 FEBS 4145

6 Lapatto R, Nalini V, Bax B, Driessen H, Lindley PF,
Blundell TL & Slingsby C (1991) High resolution struc-
ture of an oligomeric eye lens beta-crystallin. Loops,
arches, linkers and interfaces in beta B2 dimer com-
pared to a monomeric gamma-crystallin. J Mol Biol
222, 1067–1083.
7 Hejtmancik JF, Wingfield PT & Sergeev YV (2004)
Beta-crystallin association. Exp Eye Res 79, 377–383.
8 Wistow G (1990) Evolution of a protein superfamily: rela-
tionships between vertebrate lens crystallins and micro-
organism dormancy proteins. J Mol Evol 30, 140–145.
9 van Rens GL, de Jong WW & Bloemendal H (1992) A
superfamily in the mammalian eye lens: the
beta ⁄ gamma-crystallins. Mol Biol Rep 16, 1–10.
10 Jaenicke R & Slingsby C (2001) Lens crystallins and
their microbial homologs: structure, stability, and func-
tion. Crit Rev Biochem Mol Biol 36, 435–499.
11 Wistow G, Summers L & Blundell T (1985) Myxococcus
xanthus spore coat protein S may have a similar struc-
ture to vertebrate lens beta gamma-crystallins. Nature
315, 771–773.
12 Bagby S, Harvey TS, Eagle SG, Inouye S & Ikura M
(1994) NMR-derived three-dimensional solution struc-
ture of protein S complexed with calcium. Structure 2,
107–122.
13 Rosinke B, Renner C, Mayr EM, Jaenicke R & Holak
TA (1997) Ca
2+
-loaded spherulin 3a from Physarum
polycephalum adopts the prototype gamma-crystallin

fold in aqueous solution. J Mol Biol 271, 645–655.
14 Rajini B & Sharma Y (2006) Domain–domain interac-
tion and calcium-binding properties of a typical single
bc-crystallin domain of the protein absent in melanoma
1 (AIM1). Calcium Binding Proteins 1, 96–101.
15 Rajini B, Graham C, Wistow G & Sharma Y (2003)
Stability, homodimerization, and calcium-binding prop-
erties of a single, variant betagamma-crystallin domain
of the protein absent in melanoma 1 (AIM1). Biochem-
istry 42, 4552–4559.
16 Krasko A, Muller IM & Muller WE (1997) Evolution-
ary relationships of the metazoan beta gamma-crystal-
lins, including that from the marine sponge Geodia
cydonium. Proc R Soc Lond B Biol Sci 264, 1077–1084.
17 Shimeld SM, Purkiss AG, Dirks RP, Bateman OA,
Slingsby C & Lubsen NH (2005) Urochordate betagam-
ma-crystallin and the evolutionary origin of the verteb-
rate eye lens. Curr Biol 15, 1684–1689.
18 Jobby MK & Sharma Y (2005) Calcium-binding crystal-
lins from Yersinia pestis. Characterization of two single
beta gamma-crystallin domains of a putative exported
protein. J Biol Chem 280, 1209–1216.
19 Bowman GR, Smith DG, Michael Siu KW, Pearlman
RE & Turkewitz AP (2005) Genomic and proteomic
evidence for a second family of dense core granule cargo
proteins in Tetrahymena thermophila. J Eukaryot
Microbiol 52, 291–297.
20 Rajini B, Shridas P, Sundari CS, Muralidhar D,
Chandani S, Thomas F & Sharma Y (2001) Calcium
binding properties of gamma-crystallin: calcium ion

binds at the Greek key beta gamma-crystallin fold.
J Biol Chem 276, 38464–38471.
21 Duprey KM, Robinson KM, Wang Y, Taube JR &
Duncan MK (2007) Subfertility in mice harboring a
mutation in betaB2-crystallin. Mol Vis 13, 366–373.
22 Giancola C, Pizzo E, Di Maro A, Cubellis MV &
D’Alessio G (2005) Preparation and characterization of
geodin. A betagamma-crystallin-type protein from a
sponge. FEBS J 272, 1023–1035.
23 Clout NJ, Kretschmar M, Jaenicke R & Slingsby C
(2001) Crystal structure of the calcium-loaded spherulin
3a dimer sheds light on the evolution of the eye lens
betagamma-crystallin domain fold. Structure (Camb) 9,
115–124.
24 Sharma Y, Rao CM, Narasu ML, Rao SC, Somasunda-
ram T, Gopalakrishna A & Balasubramanian D (1989)
Calcium ion binding to d- and to b-crystallins. Presence
of EF-hand motif in d-crystallin that aids in calcium
binding. J Biol Chem 264, 12794–12799.
25 Sharma Y & Balasubramanian D (1996) Calcium bind-
ing properties of beta-crystallins. Ophthalmic Res 28
(Suppl. 1), 44–47.
26 Wenk M, Baumgartner R, Holak TA, Huber R,
Jaenicke R & Mayr EM (1999) The domains of
protein S from Myxococcus xanthus: structure, stability
and interactions. J Mol Biol 286, 1533–1545.
27 Bax B, Lapatto R, Nalini V, Driessen H, Lindley PF,
Mahadevan D, Blundell TL & Slingsby C (1990) X-ray
analysis of beta B2-crystallin and evolution of
oligomeric lens proteins. Nature

347, 776–780.
28 Hejtmancik JF, Wingfield PT, Chambers C, Russell P,
Chen HC, Sergeev YV & Hope JN (1997) Association
properties of betaB2- and betaA3-crystallin: ability to
form dimers. Protein Eng 10, 1347–1352.
29 Caday CG & Steiner RF (1985) The interaction of
calmodulin with the carbocyanine dye (Stains-all). J Biol
Chem 260, 5985–5990.
30 Sharma Y, Rao CM, Rao SC, Krishna AG,
Somasundaram T & Balasubramanian D (1989)
Binding site conformation dictates the color of the dye
stains-all. A study of the binding of this dye to the
eye lens proteins crystallins. J Biol Chem 264,
20923–20927.
31 Horrocks WD Jr (1982) Lanthanide ion probes of
biomolecular structure. Adv Inorg Biochem 4, 201–261.
32 Horrocks WD Jr (1993) Luminescence spectroscopy.
Meth Enzymol 226, 495–538.
33 Maruyama K, Mikawa T & Ebashi S (1984) Detection
of calcium binding proteins by
45
Ca autoradiography
on nitrocellulose membrane after sodium dodecyl
sulfate gel electrophoresis. J Biochem (Tokyo) 95,
511–519.
All lens b-crystallins are calcium-binding proteins M. K. Jobby and Y. Sharma
4146 FEBS Journal 274 (2007) 4135–4147 ª 2007 The Authors Journal compilation ª 2007 FEBS
34 Bohm G, Muhr R & Jaenicke R (1992) Quantitative
analysis of protein far UV circular dichroism spectra by
neural networks. Protein Eng 5, 191–195.

35 Slingsby C & Bateman OA (1990) Quaternary
interactions in eye lens beta-crystallins: basic and acidic
subunits of beta-crystallins favor heterologous
association. Biochemistry 29, 6592–6599.
36 Clout NJ, Basak A, Wieligmann K, Bateman OA,
Jaenicke R & Slingsby C (2000) The N-terminal domain
of betaB2-crystallin resembles the putative ancestral
homodimer. J Mol Biol 304, 253–257.
37 Norledge BV, Trinkl S, Jaenicke R & Slingsby C (1997)
The X-ray structure of a mutant eye lens beta B2-crys-
tallin with truncated sequence extensions. Protein Sci 6,
1612–1620.
38 Beard NA, Laver DR & Dulhunty AF (2004)
Calsequestrin and the calcium release channel of skeletal
and cardiac muscle. Prog Biophys Mol Biol 85, 33–69.
39 Gelebart P, Opas M & Michalak M (2005) Calreticulin,
aCa
2+
-binding chaperone of the endoplasmic reticu-
lum. Int J Biochem Cell Biol 37, 260–266.
40 Liang MP, Banatao DR, Klein TE, Brutlag DL &
Altman RB (2003) WebFEATURE: an interactive web
tool for identifying and visualizing functional sites on
macromolecular structures. Nucleic Acids Res 31,
3324–3327.
41 Van Montfort RL, Bateman OA, Lubsen NH &
Slingsby C (2003) Crystal structure of truncated human
betaB1-crystallin. Protein Sci 12, 2606–2612.
42 Blundell T, Lindley P, Miller L, Moss D, Slingsby C,
Tickle I, Turnell B & Wistow G (1981) The molecular

structure and stability of the eye lens: x-ray analysis of
gamma-crystallin II. Nature 289, 771–777.
43 Delamere NA & Paterson CA (1981) Hypocalcaemic
cataract. In Mechanisms of Cataract Formation in the
Human Lens (Duncan G, ed.), pp. 219–236. Academic
Press, New York, NY.
44 Duncan G & Jacob TJ (1984) Calcium and the physiol-
ogy of cataract. Ciba Found Symp 106, 132–152.
45 Vanmarle J, Jonges R, Vrensen GF & Dewolf A (1997)
Calcium and its localization in human lens fibres: an
electron tomographic study. Exp Eye Res 65, 83–88.
46 Dirks RP, van Genesen ST, KrUse JJ, Jorissen L &
Lubsen NH (1998) Extralenticular expression of
the rodent betaB2-crystallin gene. Exp Eye Res 66,
267–269.
47 Brunekreef GA, van Genesen ST, Destree OH &
Lubsen NH (1997) Extralenticular expression of
Xenopus laevis alpha-, beta-, and gamma-crystallin
genes. Invest Ophthalmol Vis Sci 38, 2764–2771.
48 Coop A, Wiesmann KE & Crabbe MJ (1998) Translo-
cation of beta-crystallin in neural cells in response to
stress. FEBS Lett 431, 319–321.
49 Jobby MK & Sharma Y (2003) Rapid purification of
recombinant betaB2-crystallin using hydrophobic inter-
action chromatography. Protein Expr Purif 28, 158–164.
50 Werten PJ, Carver JA, Jaenicke R & de Jong WW (1996)
The elusive role of the N-terminal extension of beta
A3- and beta A1-crystallin. Protein Eng 9, 1021–1028.
51 Marley J, Lu M & Bracken C (2001) A method for effi-
cient isotopic labeling of recombinant proteins. J Biomol

NMR 20, 71–75.
Supplementary material
The following supplementary material is available
online:
Fig. S1. A sample of recombinant bB2- and bA3-crys-
tallin resolved on 15% SDS ⁄ PAGE to determine the
purity.
Fig. S2. The putative calcium-binding sites visualized
on the crystal structure of bB2-crystallin.
Fig. S3. ITC thermogram of an inactive preparation
of bB2-crystallin.
This material is available as part of the online article
from
Please note: Blackwell Publishing is not responsible
for the content or functionality of any supplementary
materials supplied by the authors. Any queries (other
than missing material) should be directed to the corres-
ponding author for the article.
M. K. Jobby and Y. Sharma All lens b-crystallins are calcium-binding proteins
FEBS Journal 274 (2007) 4135–4147 ª 2007 The Authors Journal compilation ª 2007 FEBS 4147